*

1 191s t ELECTROCHEMICAL SOCIETY MEETING, MONTREAL, QUEBEC, CANADA (May 4-9, 1997)

BNL-646 6 2

UNDERPOTENTIAL DEPOSITION OF Ag ADLAYERS ON Pt(ll1): STRUCTURES AND DETERMINATION OF 0 2 ADSORPTION ON Pt(ll1)

CohF- 77 0 s /a- - N.S. MarinkoviC, J.X. Wang and R.R. AdZiC

Chemical Sciences Division, Department of Applied Science Brookhaven National Laboratory, Upton NY 1 1973

ABSTRACT

The structure of Ag adlayers deposited at underpotentials in sulfuric acid on Pt (lll), and the inhibition of Q reduction they cause, have been studied using grazing incident angle x-ray diaaction measurements, as well as linear sweep voltammetry and in situ FIlR spectroscopy. Ag forms a hexagonal incommensurate bilayer, with two mutually commensurate monolayers. It is aligned with the Pt(ll1) substrate, although slightly expanded. The fist monolayer has a commensurate (1x1) structure. A second layer causes a restructuring of the first monolayer. Deposition of each monolayer is associated with one voltammetry peak A complete inhibition of 02 reduction on Pt(l11) has been observed upon deposition of both, Ag monolayer and bilayer. Analysis of the inhibition of 0 2 reduction as a knction of the Ag coverage shows that during reduction adsorbs in a bridge configuration on Pt( 1 1 1).

INTRODUCTION

The underpotential deposition (UPD) of Ag on Pt has recently attracted a coniiderable interest (1,2,3,4,5,6,7). Despite new insights into the behavior of this system, it is not yet filly understood. A formation of two monolayers at underpotentials is interesting feature of this system. In addition, unlike many other metal adatoms that have small or negligible effeds on a reduction on Pt, Ag causes a complete inhibition of this reaction (for review see references 8). In this work, surface X-ray Scattering (SXS) and Fourier transform inkred spectroscopy (FTIR) techniques were used to characterize the UPD of Ag on Pt(ll1) electrode in sulfuric acid solution. SXS technique is particularly suitable for CWerization of two layers of Ag before bulk deposition, which is not possible by scanning probe techniques. Structural information obtained by SXS technique was used to analyze the inhibition of 02 reduction produced by the Ag adatom. This inhibition provides a possibility to use these adatoms as a probe of adsorption reactions on electrode surfaces, with the position of -4g adatoms on Pt(ll1) verified by SXS measurements during the course of reaction. The type of adsorption on the Pt electrode surfaces has been a long- standing question of oxygen electrocatalysis. No distinction between the three models of 02 adsorption, viz., Paling (on-top, single bond), Griffths (on-top, double bond) and “bridge”, could have been made so far.

EXPERIMENTAL

Platinum disks (3 mm by 10 mm and 6mm x 6mm) were aligned, polished and annealed as previously reported. The electrochemical-x-ray scattering cell was described in reference (9). The solutions were prepared fiom AglzS04 or AgNa (Aldrich) and HzS04 (Seastar) and Milipore QC W Plus water (Milipore Inc.). At a potential slightly positive of Ag deposition potential, the cell was deflated which leaves a thin lop-thick capillary electrolyte film between the Pt(ll1) face and the Prolene x-ray window. A reversible hydrogen electrode in Ag-free solutions was used as a reference electrode. The potentials are given with respect to standard hydrogen electrode (S.H.E.). Rotating hanging meniscus electrode measurements were carried out with a crystal 6mm in diametery 6mm long. The rotator (Pine Instrument Co.) was fitted with a collet crystal holder which is similar to the one described earlier (IO).

SXS measurements were carried out at the National Synchrotron Light Source (NSLS) at beam line X22B with 1 = 1.54k A full description of the electrochemical SXS technique has been presented elsewhere (9). For Pt(ll1) it is convenient to use a hexagonal coordinate system in which Q = (a*,b*,c*)x &l,yL), where a* = b* = 4d3a = 2.614A-', c* = 2 d 6 a = 0.924&', and a = 2.775 A The in-plane diffraction measurements were carried out in the (H,K) plane with L=0.2 conesponding to a grazing angle of 1.25 degree.

In sisitu spectroeletrochemical measurements were done in a cell described earlier (1 1). 40% scans are coadded with parallel polarized light at sample and reference potentials using potential-step technique. Spectra are presented in -AlUR form, so that positive-going bands represent a gain of a particular species.

RESULTS AND DISCUSSION

Voltammetry of the UPD of Ag on Pt(ll1) in H2S04 Solutions --- Fig. 1 shows a linear potential voltammetry m e for the UPD of Ag on Pt(l11) in

0.05M H2S04 Containing 1mM Ag+. Two pronounced peaks in the anodic sweep OCCUT at potentials 0.73 and 1.08V. The corresponding peaks in cathodic direction OCCUT at 0.68 and 1.06V. These peaks are usually ascribed to a deposition of two 9 monolayers. The charge calculated from the voltammetry w e c~rrespnds to 249pWcm for the peak at 1.1V and 261pcm' for the peak at 0.68V. These charges are close to the value of 240pC/cm2 which corresponds to a 1 1 1 monolayer of Ag. The x-ray diffraction measurements, presented below, also show a formation of a Ag bilayer with one monolayer associated with each

Previous vottammetry studies pointed out a problem of the instability of Pt(l11) at high anodic potentials in the UPD of Ag (12). With the well-ordered Pt(l11) in sulfuric acid sobions this does not appear as a problem since a strong adsorption of bisulfates prevents the oxidation of Pt at potentials of Ag dissolution.

peak

DISCLAIMER

This report was irepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, make any warranty, express or implied, or assumes any legal liabili- ty or responsibility for the accuracy, completeness, or usefulness of any information, appa- ratus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commeraal product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessar- ily state or reflect those of the United States Government or any agency thereof.

SXS Measurements

Grazing incident angle x-ray difiiaction measurements were performed in order to determine the Surface structure of electrodeposited silver adlayers in the absence and in the presence of 0 2 Fig. 2 shows typical in-plane diffiction scans along K direction at two different potentials. At 0.85 V, there is a diffiction peak only at the integer position (0, 1, 0.3). This peak is very weak compared to that measured at 1.05 V with no Ag adatoms on the pt(ll1) sdace. The electron density of Ag is nearly half that of Pc, and the intensity of the (0, 1,0.3) reflection, which is close to the anti-Bragg position, (0, 1,O.S) along the (0, 1) non-specular rod, is expected to decrease upon formation of a pseudomorphic Ag monolayer. This is indeed observed experimentally which means that Ag forms a commensurate (1x1) monolayer on Pt(l11) at potentials between the two voltammetry peaks. At 0.65V, negative of the second peak, additional difEaction peaks were found at (0, 0.940), (0,0.940), (0.940,0.940), (0, 1.880) and (1.880,O). This diffraction features indicate the formation of an aligned hexagonal bilayer with the lattice constant of 2.952 k 0.003 A which is larger than that of bulk Ag. The question of the unusual expansion of the Ag incommensurate bilayer on pt( 1 1 1) will be discussed elsewhere (13). A second layer causes a restructuring of the first monolayer. From the observed in-plane diffiction pattern, and the intensity profile of the surface rods, it was found that the two adlayers are mutually commensurate (13). The Ag bilayer is stable after removing Ag+ ions from the solution. The diffiction pattern and the real space model derived from it are shown in Fig. 3.

Oxygen Reduction on Pt( 1 1 1) / Ag

Structural information on the foreign metal adlayers provide a possibility for a more complete study of their electrocatalytic effects in various reactions and their use as a probe of adsorption processes at electrode surfices. To obtain such information requires monitoring the adlayer structure during the course of the reaction. It can be achieved by utilizing in situ SXS technique (14,15). Some limitations imposed by the cell construction imposed on the use of this technique in such measurements were discussed elsewhere (14). Limitation for 02 reduction appear to be minor since it can be brought by diffision through a thin Prolene film covering the cell perpendicular to the Surface being investigated. There is, however, a problem of the current distribution. The current at the edge of the electrode is certainly larger than at the center of the front face of the crystal when counter electrode is around the crystal. This apparently does not cause a problem at small current densities which is the case with this reaction. It has been shown that the structure of the Strudme of TI adlayer vanishes with 0 2 reduction, but recovers when 0 2 is replaced by NZ (14,15). This shows that the reaction occurs predominantly at the front surface.

Fig. 4 shows 02 reduction on Pt(ll1) obtained with a hanging meniscus rotating disk electrode in the absence and in the presence of the Ag adlayers. The m e for Pt( 1 1 1) is in agreement with the literature data (16,17,18). The Ag monolayer causes a complete

inhibition of 02 reduction The reduction is also inhibited by the Ag bilayer. This may not be surprising, given very negative potentials of the reaction on bulk silver. The reaction on a Ag monolayer in solution containing no Ag+ ions shows that the 0 2 reduction starts at potential of 0.25Vy which is more negative than on bulk Ag( 1 1 1). a reduction undergoes a four-electron reduction on bulk Ag and exhibits a small structural dependence (19). In acid solutions 02 reduction starts at -0.3V, which is considerably more negative than on R.

Fig 5 shows the effect of Qz on the intensity of the difiiaction peak at (0,0.94) position in the presence of Ag+ in solution A comparison with the measurement in the absence of 02 at 0.W shows a negligible decrease of the diffiction intensity. In the absence of Ag+ in solution, 02 reduction occurs at 0.2 V and causes a considerable decrease of the diffiction intensity. The peak position is unchanged which means that the Ag coverage and the adlayer structure remains unchanged in 02 reduction (Fig. 4). This decrease in the diffiction intensity is similar to the decrease observed during Qz reduction on the hexagonal TI adlayer on Au (1 11) (14). This was ascribed to a decrease of the two- dimensional order of the adlayer, which decreases the difiaction intensity. These measurements indicate that the structure of Ag submonolayers is not changed during the course of 02 reduction.

In situ FTIR spectroscopy measurements

Another possible explanation of the inhibition of 02 reduction is a strong adsorption of bisulfate anions on Ag adlayer. A strong adsorption of tetrahedral bisulfate on Pt(ll1) is well established (20). Fig. 6 displays Fourier Transform Mared Spectroscopy (FTIR) spectra obtained in situ 5OmV apart, starting at 0.7V and using the reference potential of 0.68V, the potential of a bilayer formation. Bisulfate is adsorbed on the monolayer as indicated by the potential-dependent band at - 1200 cm-'. Upon dissc?!ution of Ag, a band for bisulfate adsorption on Pt becomes pronounced. A solution phase sulfate band, caused by anion migration and changes of the sulfatehisulfate equilibrium due to bisulfate adsorption, is seen at 1100 cm". A presence of a strong bipolar surface bisulfate band in the potential region 0.7V I E I 0.95 V shows that the bisulfhte adsorption on the bilayer is smaller than at the Ag monolayer. Coadsorption of sulfatelbisulfate was observed for Ag monolayer on Au(ll1) electrode (21). In HF and triflourmethane sulfonic acids, which have no specifically adsorbed anions inhibition of reduction indicates that anion adsorption is not the cause. This is supported by a pronounced inhibition caused by small and intermediate Ag coverage which cannot dsorb tetrahedral anions.

Model of the site - blocking for 02 adsorption by Ag adaioms

The essential first step in electroeztalytic 02 reduction is its adsorption on the electrode surfacey M, viz.,

O2 + M = M-@(ads) 111

The absence of this step would preclude 0 2 reduction, unless the reaction can occur by an outer sphere charge transfer. The lack of 0 2 reduction currents on the A@( 1 1 1) surface shows that the outer-sphere charge transfer is not operative at the potentials investigated. It is assumed that the 02 adsorbs in a bridge configuration at Pt since it is conducive to a splitting of the 02 molecule and ensuing four-electron reduction (22). There is no spectroscopic method to provide in sihr information on 02 adsorption. The inhibition by Ag adatoms can provide some information on this question of a considerable importan* for oxygen electrocatalysis. The use of metal adatoms as a probe of adsorption processes was demonstrated on several examples ( 8,23).

The measurements of the inhibition of 0 2 reduction by Ag adatoms were carried out with Pt( 1 1 1) as a fbnction of the Ag coverage in the absence of Ag' ions in solution. Ag was adsorbed by exposing the Pt(ll1) crystal to N2-saturated solutions containing various concentrations of Ag' for short times. The Ag coverages were determined in a separate cell and excess of Ag desorbed. Figs. 7 shows 02 reduction as a fbnction of Ag coverages. A very large decrease of the 02 reduction current at small Ag coverages is observed.

In order to determine type of 02 adsorption on the Pt(ll1) electrode surface, a model of the site-exclusion by Ag atoms is needed. A pronounced inhibition observed indicates that at least the near-neighbors R atoms must be affected by Ag adatoms. Fig. 8 shows the models for the site blocking by one Ag atom for adsorption of @ in a bridge (a) and in an on-top (b) configuration. Ag atom adsorbed in the hollow site blocks five bridge sites for 02 adsorption. Only three sites are blocked for on-top adsorption. Conversely, five three-fold symmetry sites for Ag adsorption have to be unoccupied to form ensemble of Pt atoms for adsorption of one 0 2 molecule at the bridge site. Likewise, three hollow sites have to be free from Ag for adsorption of 02 at the top site. Protopopoff and Marcus (24) analyzed a range of possible effects of adsorbed sulfixr on H2 adsorption and evolution using similar approach.

This system can be treated quantitatively by using simple statistical treatment. Ifn is the number of 0 2 sites blocked by one adsorbed Ag atom, an 02 molecule can adsorb at a particular site only if n Ag sites are not occupied by Ag adatoms. The number of distinguishable ways in which q indistinguishable single particles are arranged on a lattice of N equivalent sites is given by (25,26),

W(q&) = ( N ) = N!/(N - q)!q! ( 2 ) 4

The ensemble of Pt sites, which has to be unoccupied by Ag in order for 0 2 to adsorb in a bridge, position occurs ( ) times since on the remaining ( N - 5) hollow sites, q

indistinguishable Ag atoms must be arranged. The probability of having such a structure

N - 5 4

is

P = ( N - J ) / ( N ) Q 9

131

By defining the Ag coverage by 8 = q / N, and assuming N is very large, the equation results into a simple expression P = (1 - e) 5 , or at low coverages, P = ( 1 - 58 1. The slope of the initial part of this hnction at small mverages is -5. For the case of the on-top ch adsorption, the corresponding hction will have the (1 - e>' and the slope for the inhibition at low Ag coverage would be -3. The plot of the normalized current for Q reduction on Pt(l11) as a knction of the Ag coverage at four different potentials is given in Fig. 9. The slope of the curve defined by points at small coverages is approximately -5. The calculated slope is given by dashed line. Therefore, the data in Fig. 9 provide first evidence, albeit indirect, for the bridge adsorption of 02 on Pt(ll1) during the reduction in acid solutions. Gas phase adsorption data at 15OK indicate a presence of a bridge and of on-top configurations (27). Chan et al. (28), using molecular orbital calculations, concluded that 02 chemisorption on Pt( 1 1 1) is favored at the two-fold site.

CONCLUSIONS

The UPD of Ag on Pt( 1 1 1) represents an interesting UPD system which involves a bilayer formation before bulk Ag deposition. Ag forms a hexagonal incommensurate bilayer at underpotentials while the first monolayer has a commensurate (1x1) structure. Ag adatoms at coverages 5 1 reside in the three-fold symmetry hollow sites. The mutual commensurability of the two adlayers constituting a bilayer, and a small bilayer expansion with respect to the Pt(ll1) substrate, is important information about metal adlayers that can be obtained only with the SXS. The data presented above illustrate the usefihess of the st~~ctural infomation in studies of electrocatalytic reactions. In addition, they confirm the unique possibility of the SXS technique for determining the structure of electrode surfaces during the course of electrochemical reactions. The evidence was found for the 9 adsorb in a bridge configuration on F%(lll), derived from the data on the inhibition of 02 reduction by silver adatoms.

ACKNOWLEDGEMENTS

The authors thank S. Feldberg and B. Ocko for usefbl discussions. This research was pedormed under the auspices of the US Department of Energy, Division of Chemical Sciences and Materials Sciences Division, Ofice of Basic Energy Sciences under Contract NO. DE-ACO2-76CHOOO16.

REFEXENCES

1. G.W. Tindal and S. Bruckenstein, Electrochim. Acta, 16,245 (1971). 2. S. Stucki, J. Electroanal. Chem., 80,375 (1977). 3. T. Chierchie, C. Mayer, K. Juttner, and W.J. Lorenz, J. Electroanal. Chem., 191,401

4. P.B. Costa, J. Canullo, D.V. Moll, RC. Salvarezza, M.C. Giordanno and AJ. Arvia, J.

5. F.EI O m , R Durand and RFaure, J. Electroanal. Chem., 160,385 (1984). 6. N. Kamizuka and K. Itaya, Farad. Disc. 94, 117 (1992). 7. A. Vashkevich, M. Rosenblum and E. Gileadi, J. Electroanal. Chem. 383, 167 (1995). 8 RR.Adzic, in Advances in Electrochemistry and Electrochemical Engineering, vol. 13,

9. J.Wang, B.M. Ocko, AJ. Davenport and H.S. Isaacs, Phys. Rev. B 46,10321 (1992);

10. B.D. Cahan and H.M.Villulas, J. Electroanal. Chem., 307,263 (1991). 11 P.W. Faguy and N.S. Marinkovic, Anal. Chem., 67,2791(1995). 12. J.F. Rodriguez, D.L. Taylor and H.D. Abruna, Electrochim. Acta, 38,235 (1993). 13. J.X.Wang, N.S. Marinkovic, RR. Adzic, B.M. Ocko,. Phys. Rev. Lett. submitted for

14. RR. Adzic, J.X. Wang, B.M. Ocko, Electrochim. Acta, 40, (1995) 83. 15. RR. Adzic, J.X. Wang, in “Oxygen Electrochemistry”, R.R. Adzic, F.C. Anson, K.

(1 985).

Electroanal. Chem., 244,261 (1988).

H. Gerischer and C. Tobias, Editors, p. 159 John Wiley, New York, 1984.

B.M Ocko, J.Wang, AJ. Davenport and H.S. Isaacs, Phys. Rev. Lett. 65,1466 (1990).

publication.

Kinoshita, Editors, PV 95-26, p. 61 The Electrochemical Society Proceedings Series, Pennington NJ (1996).

16 N.M. Markovic, H. Geistarger and P.N.Ross, J. Phys. Chem., 99,341 1 (1995). 17 F. El Kadii,RFaure, R. Durant, J.Electroanal. Chem., 301,177 (1991). 18 N.M. Markovic, R.R. Adzic, B.D. Cahan and E.B. Yeager, J. Electroanal. Chem.,

377,249 (1 994). 19. See e.g. R.R Adzic, in Modern Aspects of Electrochemistry, vo1.21, p. 163, R.E.

White, J.O’M. Bockris and B.E. Conway Editors, Plenum Press (1990) New York. 20. P.W. Faguy, N.M. Markovic, R.R. Adzic, C.A. Fierro, E. B.Yeager, J. Electroanal.

Chem., 289,245 (1990); P.W.Faguy, N.S. Marinkovic, R.R. Adzic, J. Electroanal. Chem., 407,209 (1996).

21. P. Mrozek, Y.-E. Sung, A. Wieckowski, Surf. Sci., 33544 (1995). 22 E.Yeager, M.Razaq, D. Gervasio, A. Razaq, D. Tryk, J. Serb. Chem. SOC., 57,819

23. RR Adzic, B.Z. Niiolic, F. Feddrix and E.Yeager, 3.Electroanal. Chem., 341,287 (1 992).

(1 992).

24 E. Protopopoff, P. Marcus, J. Chim. Phys., 88, 1423 (1991). 25. EMyazaki and I. Yasumori, J. Math. Phys., 18,215 (1977). 26. G.A Martin, in Metal-Support and Metal-Additives Effects in Catalysis, B.Imelik et

27. H. Steininger, S. LehwaId and W. Ibach, Sd. Sci., 123, 1 (1982). 28. A. W.E. Chan, R. Haffman and W. Ho, Langmuir, 8,1111 (1992).

al. Editors, Elsevier, Amsterdam (1982).

I I

30 2 mV/s -

I Dissolution i ___t

I'

t bulk 2nd layer 1st layer -30 -

0.6 0.7 0.8 0.9 1 1.1 1.2 . E / mV vs. R.H.E.

20 .

3i 10 \

E O 5 -10

-20

E 0

Fig. 1 Voltammetry curve for the UPD of Ag on Pt(ll1) in 0.05m &SO4 with 1mM Ag+. Sweep rate ZmVIs.

n W v 1.5

5 1.125

* 0.75 'z v X Fig. 2 In-plane diffractioc .- ,x 0.375 scans along the (O,K,0.3) axis c. c W from on Pt(ll1) in O.05M 5 0 &SO4 with 1mM Ag+. Solid

002 0.94 0.96 0.98 1 1.02 lines are the fits to a Lorenzian

5 0 0

v)

K (units of b') line shape.

(a) Diffraction Pattern (b) Real Space Model

f J - $ 8 (0.1) -

(1 8 0 )

Fig. 3 In-plane diffraction Q Q pattern tiom Pt( 11 1) in 0.05m

H2S04 with lmh4 Ag+ at 0.65V 0 0 and the corresponding real

space model of the Ag bilayer.

*) ec. I

0

3c. nn +

4 \ 4d

5 2 L. J 0 I

90 30 '

100.0

270.0

360.0-

450.0-

540.0-

630.0-

720.0-

1 i I

810.0! I I I I a I I , . I I I I ~, +0.05 +0.20 +0.35 4.5Q +0.65 4.80 +0.95

PotentioI/V

Fig. 4 02 reduction on a rotating hanging meniscus Pt( 1 1 1) with and without Ag monolayer.900 rpm; SOmV/s.

8000

6000

4000

h 2000 u 0, s o - c

In-plane diffraction Fig. 5 scans along the (O,K,O.3) axis fiom on p t ( l l 1 ) at 2000

4000 1 n l ] indicated potentiis in the " I absence (A) and in the

presence of 0 2 (0). 0.92 0.93 0.94 0.95 0.96 0.97 0.98

K (units of b')

- AR/R

0.5%

1200 1100 1000

wavenumbers / c m - I

Fig. 6 SNIFTIR spectra from Pt(ll1) in 0.05M H2S04 with 1mM Ag+ at sample potentials from 0.7 to 1.15V increment4 by 50mV. Reference potential 0.68V; p- polarization.

-100 -

a

V 400 500 600 700 800 900

E / mV vs. R.H.E. I000

Fig. 7 0 2 reduction on Pt( 1 1 1) in 0.05M H2S04 as a hnction of Ag coverage.

,

0 2 “BRIDGE” 0 2 ON-TOP

Fig. 8 Models of the site-exclusion by Ag atom for the “bridge’ and on-top 02 adsorption.

1.0

0.8

0.6

- 0.4 \ -0

0.2

0.0

A

0 rn

0 0.8V D 0.775V 7 0.75V A 0.7V

I 1 , 1 , , , 1 ,

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Fig. 9 Normalized 0 2 reduction current as a fhnction of Ag coverage at 4 potentials. Solid line: extrapolation at low coverages; dashed line: extrapolation of the calculated curve.